[MPlayer-users] Re: Shared Postprocessing? - MPlayer/MEncoder last CVS compiling problem

D Richard Felker III dalias at aerifal.cx
Mon Nov 4 00:25:02 CET 2002


On Sun, Nov 03, 2002 at 10:25:32PM +0100, Michael Niedermayer wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> Hi
> 
> On Sunday 03 November 2002 19:51, Andrzej Nikitorowicz wrote:
> > [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> >
> > On Sat, 2 Nov 2002:
> > > "," -> "/" so hb,vb -> hb/vb
> > >
> > > thats because "," is used as seperator of the filters for -vop
> > >
> > > Michael
> >
> > Thank you very, very much. Mplayer is compiling, postprocessing works
> > great.
> >
> > (...)
> > Opening Video Filter:
> > [pp=hb:90:5/vb:90:5/h1/v1/lb/li/ci/md/fq:3/tn:64:128:256]
> > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> > Is this O.K?
> the format is correct but i doubt that the combination of filters is a good 
> choice 

To elaborate on Michael's comment...

You should only use one or the other of hb and h1, not both. Same goes
for vb and v1. From what I understand, hb and vb are better tested,
but h1 and v1 may be a bit faster.

All of lb, li, ci, and md are deinterlacing options. You should rarely
(if ever) use more than one at a time. Also, they will degrade image
quality, so they should ONLY be used if the source movie is actually
interlaced and needs to be fixed.

As for fq, it should probably never be used, unless perhaps you're
using windows binary codecs that don't export a quantizer table. The
deblocking filters will automatically get the appropriate quantizers
on a per-block basis if you use libavcodec or libmpeg2 for decoding.

Rich




More information about the MPlayer-users mailing list